The same question, asked five times
Research, briefs and analyses done once in the Project are visible to everyone. The second person builds on the first instead of starting over.
Team projects with AI
Team projects with AI: one shared context, every model, one invoice.
When a team uses AI separately, the team's knowledge fragments into personal chat histories. Two people research the same competitor. A new hire re-asks what the AI told someone in March. A decision is made in one person's conversation and nobody else knows. Shared Projects fix this: the team works from the same files, instructions and history, with any AI model, and the context outlives any one person.
Pro for Teams and Smart for Teams include shared Projects, chats, prompts and files, seat management and one invoice.
Where teams lose knowledge
Four leaks that shared Projects close
The new hire who is useful on day one
They read the Project: the customer, the decisions, the files, the reasoning. They ask it what they need. No week of interruptions.
Choices nobody else can find
A decision log in the Project, read by every model. The AI stops proposing what the team already rejected, and people stop relitigating it.
Context that leaves with a person
When someone goes, the Project still holds their files, prompts and reasoning. The team keeps what it paid for.
How teams use it
Patterns that work
| Team | Project per | What lives in it |
|---|---|---|
| Agencies and consultancies | Client | Brief, voice, files, deliverables, feedback, reports. See the agency workspace. |
| Product and engineering | Product or codebase | Spec, decisions, selected repo files (GitHub import where enabled), reviews, release notes. See coding with AI. |
| Marketing | Campaign or channel | Positioning, audience research, drafts, performance data, what worked. |
| Founders and leadership | The company | Plan, financial model, board updates, research. See the startup workspace. |
| Research and analysis | Topic or study | Sources, findings with locations, the open questions. See research with AI. |
Several models, one context, whole team
A shared Project means the analyst can ask Claude, the writer can ask ChatGPT and the engineer can compare both, all against the same files. Disagreements between models become team discussions with the evidence in view.
Plans and administration
What the team plans include
Pro for Teams ($25 per seat per month) includes everything in individual Pro plus centralised billing, admin seat management, shared Projects, shared chats, shared prompts and files, one invoice and priority email support. Smart for Teams ($95 per seat per month) adds higher per-seat usage and upload limits, advanced collaboration modes for the team, and dedicated onboarding for teams of ten or more. Both can be billed monthly or yearly. For SSO, audit logs, data residency, custom terms or your own model keys, Enterprise has custom pricing. Current details are on the plans page.
Content stays inside your workspace, is encrypted, and is never used to train any model. Members leave with their seat, not with the Project.
Free is the start. The latest models are the difference.
A team's decisions deserve the strongest models, on the same context.
The team plans are where shared Projects live, and they carry the latest models across every family for every seat. That combination, several strong models reading one shared context, is what turns individual AI use into a team capability: fewer duplicated efforts, faster onboarding, decisions with the evidence attached.
Newer models hold your whole Project in view instead of the last few messages, reason through trade-offs instead of picking the first plausible answer, and are wrong less often and with less confidence. For the questions on this page, that is the difference between advice that sounds right and advice you can act on.
Pro from $20 a month, cancel any time. The free plan stays free.
- Your first Project and file uploads
- The major AI families to try the workflow
- A small daily message allowance
- The latest ChatGPT, Claude, Gemini, Grok and Perplexity models, all on the same Project
- Far more messages a day, so a working session does not stop halfway
- Modes where several models draft, challenge and verify each other's answers
- Shared Projects, chats, prompts and files across the team
- Admin seat management, centralised billing, one invoice
- Higher usage and upload limits on Smart for Teams
Start here
First messages for a team Project
Add the files, the instructions and the decision log. Then start here.
OrientationWrite the one-page orientation for a new team member from this Project: what we do, current priorities, key decisions, where things are.
DuplicatesAcross the chats and files in this Project, what has been researched or analysed more than once? Consolidate it.
Decision logExtract every decision from the shared chats this month, with date and reasoning, into the decision log.
BriefDraft the campaign brief from the audience research and positioning in this Project, then flag what is still assumed.
ReviewReview this deliverable against the Project instructions and the last three approved pieces. What is inconsistent?
WeeklySummarise what changed in this Project this week: new files, decisions, open questions, for the Monday update.
Honest answers
What people ask
What does a team plan include?
Shared Projects, chats, prompts and files; centralised billing and seat management; one invoice; priority support. Smart for Teams adds higher limits, advanced collaboration modes and onboarding for larger teams. See plans.
Can we control who sees which Project?
Projects are shared within your team workspace. For finer controls, audit logs and SSO, Enterprise is the path.
Is our data kept separate from other customers?
Yes. Content stays in your workspace, is encrypted, and is never used to train any model. Enterprise adds data isolation and residency options. See trust.
What happens when someone leaves the team?
You remove their seat. The Projects, files and chats they contributed to stay with the team.
Can the team use different models?
Yes. Every seat can use any model in the plan against the same shared context, and compare them.
How do we start?
Create a team from the team page, invite members, and create the first shared Project. Billing is per seat, monthly or yearly.
Give the team one memory.
Create a shared Project for the thing you all work on, add the files, and stop answering the same question twice.
Also: the agency workspace · the startup workspace · compare AI models for a project · team page